SAFE: Synthetic Audio Forensics Evaluation Challenge
Figure 1: Synthetic Audio Forensics Evaluation Challenge Round 1 Results.Performance (circle markers) from top five teams and the detection vs. false alarm curves are shown for three tasks of increasing difficulty: detection of (1) synthetic voice audio, (2) synthetic voice audio post-processed with various compression and resampling and (3) laundered to evade detection.
Authors
- Michael Davinroy (ORCID: https://orcid.org/0009-0002-8735-1956)
- Matthew C. Stamm (ORCID: https://orcid.org/0000-0002-3986-4039)
- Laura Cassani
- Peter Bautista (ORCID: https://orcid.org/0000-0003-4809-8951)
- Jai Aslam (ORCID: https://orcid.org/0000-0002-6374-6446)
- Pranay Pherwani
- Trapeznikov Kirill (ORCID: https://orcid.org/0009-0009-7232-4542)
- Paul Cummer
Institutions
- Aptima (United States) (US)
- Systems and Technology Research (United States) (US)
- Drexel University (US)
Publication Details
- Published
- 2025-06-17
- DOI
- https://doi.org/10.1145/3733102.3736707
- Primary Topic
- Music and Audio Processing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Urmia Lake Research Institute, Urmia University